Abstract: The development of learning health systems is causing radical transformation of the environment within which the NCI pursues its mission; understanding the ethical and social implications of these changes is of paramount importance. In rapid learning systems (RLS), routinely collected patient data drive the process of discovery, which in turn becomes a natural outgrowth of clinical care. As the Institute of Medicine has noted, such systems have substantial promise for improving the quality of care and research, and ultimately the value of health care. As such systems develop, the blurring of the current distinction between clinical practice, quality of care, and research necessitates careful consideration of ethical implications. As RLSs are in their infancy, it is critical to conduct research to generate informed and considered patient perceptions of the ethical implementation of such systems, particularly regarding ways to ensure respect for patient autonomy and privacy, including best approaches for informing participants and governance of data use, in order to realize the potential benefits of these systems. Therefore, we propose an innovative study that uses cutting edge methods of deliberative democracy to generate considered and informed opinions of cancer patients, leveraging a unique opportunity to evaluate patient experiences during the roll-out of a real-world RLS. Specifically, the American Society of Clinical Oncology (ASCO) has developed a real-world oncology RLS known as CancerLinQ. CancerLinQ is being implemented in 15 vanguard practices over the next year, and the approach to patient notification/consent and data governance in this system is actively evolving. We propose an empirical investigation with two distinct approaches and aims, in collaboration with ASCO and its vanguard practices. First, we will use a deliberative democracy approach to determine the range of informed and considered individual and group opinions and recommendations of cancer patients on the optimal approach for obtaining consent and appropriate uses of information routinely collected in the course of medical care as part of a RLS that seeks to improve quality and advance research. Second, following CancerLinQ roll-out, we will survey patients experiencing the real-world implementation of this RLS in order to evaluate their knowledge and perceptions of that system. Conducting the proposed work in parallel with the development of a real-world RLS provides an opportunity to directly inform the development and implementation of a national learning system that will ultimately impact tens of thousands of patients, and it also allows for the consideration of real- life rather than purely hypothetical scenarios in ways that increase the likelihood that these investigations will yield insights that are directly applicable in other settings. The findings will have substantial relevance to the research mission of the NCI, as oncology learning systems are fundamentally altering the context for research across the spectrum of cancer causation, diagnosis, prevention, treatment, and survivorship care.